When an industrial unit fails, the decision to repair the board or swap the whole module is often made on emotion: replacement feels faster and simpler. But once you lay out component-level repair versus a module swap in numbers, the picture changes. For rare and discontinued units, repair frequently turns out to be not merely cheaper but the only workable option.
Two approaches at heart
A module swap is a brute-force method: the faulty assembly is discarded whole and a new one is fitted. It is fast but wasteful: along with the single failed element you throw out hundreds of healthy ones.
Component-level repair works differently. An engineer localizes the specific fault down to an individual component - a capacitor, transistor, driver, or chip - and replaces only that. The board stays the same, with proven compatibility and a history of running on your equipment. We write more about the philosophy of repairing instead of replacing in a separate article.
How the fault is localized
Component-level repair is possible because modern diagnostics can find a defect without a schematic. Analog signature analysis (ASA) instruments like BoardMaster compare the response of each node against a reference and flag the anomaly. We break down how this works in practice in our piece on fault localization with BoardMaster.
Thermal imaging under load adds a second layer of information: an overheated component shows up on the thermogram before it fails outright. That method has its own article on thermal scanning before repair. Together these tools reduce diagnostics to a targeted search rather than a guess-and-replace swap.
When repair wins
There are clear markers where component-level repair almost always comes out ahead:
- The unit is obsolete and no new equivalent can be bought.
- Lead time for a new module exceeds 8 weeks and the equipment cannot stay idle that long.
- Repair cost is under 40% of the price of a new unit.
- The board is unique or adapted to a specific installation, and a new module would need extra commissioning.
These rules apply especially in industries with expensive, long-lived equipment: marine automation, metro equipment, and oil and gas systems. Our long-standing customers, from metro operators to major transport enterprises, come to us with exactly these units, the ones for which replacement would mean months of downtime.
The problem of discontinued units
A distinct challenge in industrial electronics is its long service life. An installation runs for fifteen or twenty years, while its electronics goes out of production far earlier. By the time of the first serious failure the module is often already unavailable, and its successor is incompatible in connectors, firmware, or physical dimensions.
In that situation replacement stops being the simple option. You either hunt for a used module of dubious origin, or rework the footprint and surrounding circuitry to fit a new unit, which drags in redesign and a fresh round of commissioning. Component-level repair removes that question: the failed element almost always has an available equivalent, even if the board itself has long been out of production. That is why, for equipment older than ten years, repair is frequently not a choice but a necessity.
Weigh TCO, not the sticker price
The key mistake in choosing is to compare only the price of the part. The right metric is total cost of ownership (TCO), which includes:
- the price of the unit or the repair;
- shipping and customs for imported modules;
- commissioning and recalibration after installation;
- the cost of equipment downtime while waiting;
- the risk of a new revision being incompatible with the existing surrounding circuitry.
By the raw price of the part, replacement sometimes looks comparable. But as soon as weeks of downtime and commissioning enter the calculation, component-level repair pulls ahead. An hour of production line downtime is often more expensive than the repair itself, and that line is what most often flips the decision.
Reliability and warranty
A fair question: does repair lose on reliability? When standards are followed, no. Soldering is inspected against IPC-A-610, conductor and plated-through-hole repair follows IPC-7711/7721, and work is done with protection against electrostatic discharge (ESD).
After repair the board is not shipped to the customer right away. It goes through acceptance testing: burn-in under load and thermal cycling expose early failures, and parameters are logged before and after. That stage is described in detail in our article on post-repair burn-in. For harsh operating conditions a conformal coating is applied. We back the result with a warranty, precisely because it is measured rather than promised.
When to swap the module anyway
Component-level repair is not universal. Replacement is justified when the board is physically destroyed, when there is no access to documentation and components, or when functional safety requirements allow only a certified new unit. In those cases we honestly recommend replacement rather than dragging out a repair for its own sake.
But in our experience such situations are the minority. In most requests the cause comes down to a single worn-out or overheated component, and a targeted repair returns the unit to service in days and at a fraction of the cost of new.
We always provide an estimate before work begins, so the decision is made on numbers rather than emotion. If you are unsure which is better for your equipment - repair or replacement - request a diagnostics quote: we will assess the board’s repairability and cost out both options by TCO.